373

SEE ALSO SIDEBARS:

|||J||||'

'J|/| ||'J||||'

||JM/'' & ||J||||'
Engineered and Artificial
Photosynthesis
Cellulosic Ethanol
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
Bioenergy Research Needs for
Heat, Electricity, and Liquid Fuels
Alexander E. Farrell (University of CaliforniaBerkeley, USA)
Anand R. Gopal (University of CaliforniaBerkeley, USA)
Abstract
Biomass remains a key energy source for several billion people living in developing countries, and
the production of liquid biofuels for transportation is growing rapidly. However, both traditional
biomass energy and crop-based biofuels technologies have negative environmental and social
impacts. The overall research challenge for bioenergy is to develop the technologies to produce
useful products at low costs while minimizing the use of scarce resources such as arable land and
water. This requires substantial advancements in modern biomass power generation and the
success of liquid biofuel technologies that permit the use of lignocellulosic feedstocks or possibly
algae. With such technologies, biomass resources could meet a signicant fraction (over 10%) of
global energy demand. Both improved policies and technologies are needed to ensure that
bioenergy contributes signicantly to economic, social, and environmental goals.
Introduction
Biomass is the oldest fuel known to humankind and is still
widely used in developing countries, where it accounts for
about 35% of primary energy consumption, compared to just
3% in Europe and North America.
1
Further, because most of this
biomass is used very ineIfcientlywith adverse impacts on
public healththere is a signifcant, longstanding need Ior
improved bioenergy technologies in the developing world.
Recently, there has been renewed interest in the industrialized
world in the potential for bioenergy to mitigate global climate
change and for liquid biofuels to substitute for expensive
imported oil. This unusual synergy of interests has placed
biomass electricity and biofuels at the center of both much
excitement and much concern about their environmental and
socioeconomic implications.
2
In this article, we provide an overview of the most common
technologies currently used to generate heat, electricity, and
liquid fuels from biomass. Then, we describe the technical, eco-
nomic, and environmental opportunities and challenges faced
by each of these technologies. Because of the widespread nature
and adverse impact of biomass energy use in the developing
world, we frst address heat and electricity generation Irom bio-
mass combustion and gasifcation. We then turn to bioIuel pro-
duction Irom various Ieedstocks. We also include a separate
section on catalysis, as there is a strong need for research and
development in this area (see also Gates et al.s article on catal-
ysis in this issue). We conclude with a look at some scenarios
on the future of bioenergy.
Combustion
Direct combustion is the oldest known way to use biomass
for energy, and this method still accounts for more than
95% of global biomass energy production.
1
There is a wide
variation in the technology and applications of biomass
combustion.
A key biomass combustion technology is the open, three-
stone cooking fre that is used across the developing world.
Woody biomass or animal waste is burned underneath a cook-
ing pot supported by the stones. The eIfciency oI the three-
stone fre is very poor at approximately 15, and its users are
directly exposed to emissions of carbon monoxide, particulates,
nitrogen oxides, and tars.
3
The most common modern combustion technology is the
biomass-fueled electric power plant, in which the heat from
the Iurnace is exchanged with a working fuid to turn the prime
mover, which is attached to an electric generator. In some cases,
the plants operate as combined heat and power generators. The
system closely resembles coal-fred plants where the working
fuid is steam and the prime mover is a steam turbine. Signifcant
differences from coal plants are seen only in the furnace and the
gas cleaning system, iI one exists. The main pollutants in the fue
gas are ash, tars, ammonia, and nitrogen oxides.
4
Flue gas is
usually cleaned only if mandated by local air-quality regulations.
In some plants, useful heat is recovered from the exhaust gas.
In well-designed systems, combustion of biomass occurs in
a Iurnace. The hot fue gases that exit the Iurnace carry all oI
the usable thermal energy from combustion. A good furnace
must minimize heat loss and, hence, needs to be well insulated.
Various options exist for other design features. It is also possi-
ble to add biomass to other fuels, such as coal, in boilers origi-
nally designed for only one fuel.
5,6
In fxed-bed Iurnaces, biomass Iuel is placed on a fxed bed,
and air is supplied from two locations to allow for combustion
to proceed in steps. Both the biomass feed and the ash removal
are typically automated.
7
There are several types oI fxed-bed
furnaces, and these are still widely used for small-scale power
generation (1 MW).
Fluidized-bed furnaces for biomass combustion were intro-
duced about 25 years ago. In these furnaces, refractory materi-
als such as sand or limestone are suspended by air currents and
serve as the medium for heat transfer to the biomass fuel. High
air velocities fuidize the heat-transIer medium and the biomass
fuel. If the particles remain suspended within the furnace, the
Iurnace is called a bubbling fuidized bed, but iI the air veloci-
ties are high enough to carry material out of the furnace to be
recirculated aIter it, the Iurnace is called a circulating fuidized
bed. In general, because of the additional energy cost of circu-
lating air at high velocities, fxed beds are a better option Ior
small furnaces.
374
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
Combustion: Opportunities and Challenges
An estimated three billion people continue to rely on solid
fuels burned in the home to meet their basic energy needs.
8

Approximately two million deaths annually are related to
indoor air pollution from the burning of biomass.
9
Apart from
the obvious indoor emissions issue, there are other serious
impacts of biomass energy use by the poor, including time spent
(often by women and children) collecting fuel, deforestation (a
limited effect), degradation of soil quality (due to the use of
animal wastes as fuel), global warming due to products of
incomplete combustion, and costs (mostly to rural families).
In the near term, there is a great need for biomass cookstoves
that improve eIfciency and dramatically reduce indoor emis-
sions. It is easy to improve the 15 eIfciency oI the three-stone
fre, and any eIfciency gains will serve to reduce the associated
health and environmental problems. To be most effective, these
new technologies must be inexpensive to purchase and main-
tain, and they must be disseminated widely. In the longer term,
optimal solutions should enable the poor to cheaply and easily
access cooking fuel that does not damage their local environ-
ment or expose them to hazardous emissions. Such technolo-
gies could use different fuels but could also continue to be
biomass-based. Greenhouse gas (GHG) issues might also be
addressed in these new technologies, in ways that basic human
energy needs for cooking, heating, and lighting are supported
simultaneously with climate change imperatives.
Many biomass combustion power plants are in operation
throughout the world. The prime factors determining their loca-
tion and economics are fuel quality, quantity, seasonality, and
ease of access. As a result, biomass power plants exist only in
large-scale niche applications such as self-generation for paper,
lumber, or sugarcane processing. In order to make these plants
more competitive with fossil fuels, several key areas of research
need to be pursued.
In the very near term, co-combustion of biomass in coal
plants shows great promise as a low-cost solution to cut CO
2
,
sulfur, and nitrogen oxide emissions.
10
In the short term,
improvements in the eIfciency and alterations in the scale oI
combustion technology, including its ability to handle various
fuels of inconsistent heating value, can greatly help increase
capacity factors. Because the quality of biomass is much more
variable than that of most fossil fuels, improved fuel collection,
handling, and preprocessing can dramatically improve eIf-
ciency. In the long term, efforts are needed to bring about a
signifcant increase in the eIfciency oI the entire combustion
generation system. Currently, typical eIfciencies are in the 20
25% range for electricity generation from biomass.
11
It might
well be more effective and economical to focus research efforts
on generation Irom biomass gasifcation, as this technology
holds more promise Ior improvements in eIfciency, and to
phase out large-scale generation from combustion. Break-
throughs in the Iollowing specifc materials research areas can
make a substantial contribution to biomass combustion
technology:
N improved refractory materials for furnace walls that will
result in better thermal insulation and hotter fue gases, lead-
ing to increased thermal eIfciency, and
N reactor design and fuel processing methods that will result in
more complete combustion.
Gasication
Gasifcation is a process in which solid or liquid carbona-
ceous material, such as biomass, coal, or oil, reacts with air,
oxygen, and/or steam to produce a gas product called syngas or
producer gas that contains primarily CO and H
2
, along with
lesser amounts of CO
2
, CH
4
, and N
2
. This combustible mixture
oI gases is then burned to produce energy. About 7085 oI the
energy in the biomass can be transferred to a gaseous form.
12

Biomass gasifcation begins at a lower temperature than coal
gasifcation because biomass is more reactive than coal.
Biomass also contains potassium, sodium, and other alkali met-
als that can cause deposition oI liquefed ash (slagging) and
Iouling problems in conventional gasifcation equipment.
Biomass gasifcation is an old technology, but research and
development in it has stagnated because of low fossil fuel
prices. Many industrial routes for the utilization of biomass gas-
ifcation exist, such as production oI H
2
by the watergas shiIt
reaction and production oI diesel Iuel by the FischerTropsch
process.
12
Gasifers come in many designs, but all use one oI two oxi-
dizing agents: air or oxygen. The main advantage of oxygen
gasifcation is the production oI high-heating-value gas that can
be suitable for use in conventional natural gas turbines and also
for pipeline distribution if necessary. However, a substantial
penalty is incurred in terms of the cost and safety issues entailed
by an oxygen handling subsystem.
13
Hence, oxygen gasifcation
is rarely used, but its potential advantages should make it an
active area for research.
In steam gasifcation, steam is used as a reducing agent to
produce a gas of higher heating value. The steam reacts with
both methane and carbon monoxide to produce larger amounts
of hydrogen. However, the costs for the subsystems associated
with steam gasifcation are currently very high.
14
Larger scale (~1 MW) gasifers are generally used Ior power
production with an engine or, less frequently, with gas turbines.
Gas turbines require the fuel to be compressed, and technolo-
gies are being developed to gasify biomass under pressure to
eliminate the considerable energy cost of compressing the pro-
ducer gas. Currently, most pressurized gasifers are oI the fuid-
ized-bed type, but the technology is far from commercial.
7
Fixed-bed updraIt gasifers are the oldest and simplest
gasifcation technology, in which biomass Iuel is introduced
at the top of the reactor and air is introduced at the bottom. The
producer gas exits from the top of the reactor. The main advan-
tages oI an updraIt gasifer are its ability to handle Iuel with a
high moisture content, its simplicity and low cost, and its ability
to handle fuel with a high ash content. The main disadvantages
are the high tar content oI the gas and the low eIfciency oI
conversion.
In a fxed-bed downdraIt gasifer, the biomass Iuel and the
air move downward together; thus, the intermediate product
gases move through the hottest zones of the reactor. This burns
off most of the tars to produce much cleaner gas, which is the
main advantage oI downdraIt gasifers. The main disadvantages
are the need for strict moisture limits in the fuel and the high
amounts of nitrogen oxides and particulates in the product. This
design is very popular in applications up to 1 MW.
Fluidized-bed gasifers are very similar to fuidized-bed
combustion furnaces. A heat-transfer medium such as sand is
fuidized along with the biomass Iuel by high air velocities.
Like combustors, fuidized-bed gasifers are also designed to be
bubbling or circulating. Because of the high energy cost of air-
blowing subsystems, the benefts oI fuidized-bed reactors are
realized only at larger scales oI operation. Some oI the major
advantages oI fuidized-bed gasifers are similar to those oI fu-
idized-bed combustors, including high reaction rates and con-
version eIfciencies and the ability to tolerate a wide variation
in fuel types and characteristics.
The producer gas Irom a gasifer is directly used as a work-
ing fuid Ior power generation, which is essential to the eIf-
ciency advantage oI gasifcation over direct combustion, but
which also means that the gas must be free of contaminants. The
375
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
same is true of syngas used to produce biofuels, as discussed
Iurther below. Thus, gas cleaning downstream oI the gasifer is
an integral part oI any gasifcation technology. Biomass-derived
producer gas can contain up to several hundred parts per million
(ppm) of sulfur, depending on the source. Although this is far
below the levels Irom coal gasifcation, sulIur must be removed
from biomass-derived producer gas to levels approaching
1 ppm for fuel synthesis. In addition, removal of tars, alkali,
chlorine, ammonia, and particulates is required for virtually all
downstream conversion processes. The main contaminants, the
problems they cause, and the cleanup options are summarized
in Table I.
The removal of particulates and alkali metals is simple but
ineIfcient. Thus, there is much to be gained Irom the develop-
ment of materials and technology to remove these contaminants
at high temperatures.
Tars in the producer gas can either be removed or be
destroyed by cracking. Water scrubbing is the most widely used
method for the physical removal of tars. However, the removal
eIfcacy is not nearly as good Ior scrubbing as Ior cracking,
mainly because tars need to be coalesced and cooling alone is
not suIfcient to remove them Irom the gas stream. In addition,
this technique is fairly expensive and generates large amounts
of polluted water.
7
Thermal cracking avoids these problems, but
on the other hand, is less effective and also has an added energy
cost. Catalytic cracking can be very effective at temperatures of
800900C with up to 99 tar destruction, but is expensive.
13
Nitrogen compounds in biomass are also volatilized as
ammonia and hydrogen cyanide. During combustion, these can
form NO
x
. Fuel-bound nitrogen can be reduced in the following
ways: selecting biomass without much nitrogen, applying water
scrubbing, controlling combustion to reduce NO
x
formation,
and using selective catalytic reduction at the exhaust.
13
Gasication: Opportunities and Challenges
More than 70% of the energy content of most types of bio-
mass is easily volatilized in a gasifer, and when gasifcation is
combined with partial oxidation, up to 85% of the energy con-
tent can be gasifed.
15
Such a high energy recovery through gas-
ifcation is not possible Ior many other Iuels including coal.
When this beneft is combined with eIfciency gains through
combined-cycle gas turbine generation in a process known as
biomass integrated gasifcation combined-cycle (BIGCC) gen-
eration, the potential advantage oI gasifcation over combustion
is easily apparent.
In theory, gasifers can accept mixed Iuel inputs, making it
possible to use heterogeneous, diffuse wastes. However, sig-
nifcant barriers remain beIore this potential can be realized.
Biomass gasifcation is not signifcantly commercialized, so its
costs are much higher than those of conventional combustion-
based electricity generation. In addition, some serious technical
issues need to be overcome beIore biomass gasifcation can
become widespread. Foremost among these are yet-to-be-
proven treatment technologies for the producer gas to remove
contaminants that can damage internal combustion engines and
gas turbines.
In the long term, gasifcation should be able to oIIer benefts
to the vast unelectrifed areas oI the developing world that have
dispersed and low-quality Iuel. Small (50 kW), low-cost gas-
ifers that could operate reliably in rural villages, vary output
readily, and handle heteroge-
neous biomass fuel would be in
great demand. These units also
need to be rugged and prefabri-
cated for easy assembly and low-
maintenance operation in the
rural areas of the developing
world.
Anaerobic Digestion
Anaerobic digestion (AD) is
the breakdown of organic mate-
rial by a microbial population
that operates in an oxygen-free
environment. AD is of great
interest in energy production
because the main product of the
digestion process is methane, a
powerful greenhouse gas. However, AD is primarily used as a
waste management technology to eliminate pathogens, reduce
odors, prevent nutrient contamination of groundwater, produce
useful fertilizers, and eliminate releases of methane.
16
In addi-
tion, AD can produce a gas that can be used for heating or elec-
tricity generation.
Because AD was commercialized as a waste management
technology, it is substantially more mature than many other
renewable energy technologies. In the United States and Europe,
AD has primarily been employed in animal husbandry and in
wastewater treatment as a cheap, effective, and environmentally
low-impact means of dealing with waste. The product gas from
AD is not necessarily used for electricity and heat generation,
but simple retrofts are Ieasible Ior most digester systems.
Although little attention has been paid to AD in recent years,
this technology can play an important role in reducing GHG
emissions in two ways.
17
First, untreated organic waste matter
aerobically decomposes to release methane and carbon dioxide.
When waste is instead collected and directed to a digester and
used for energy production, these methane emissions are
avoided. Second, the biogas produced can be used Ior com-
bined heat and power, frequently offsetting a more GHG-inten-
sive source of power. The avoided methane and CO
2
emissions
from waste collection are, in most cases, much larger than the
emissions offset by electricity and heat production.
18
Lower
chemical fertilizer use in farms where the solid residues of AD
are spread on the feld is another source oI avoided emissions
that must be counted.
Opportunities to capitalize on these GHG emissions reduc-
tions are important for the adoption of AD technologies. In their
absence, many AD installations in the United States and Canada
do not produce power because the economics dictates against
it.
19
Further, small animal farms across the developing world are
responsible for substantial methane emissions. These farmers
could afford AD technologies if they were able to access carbon
markets that would make such investments worthwhile. Once a
5BCMF* Produced Gas Contaminants, Problems, and Cleanup Processes.
Contaminants Examples Problems Cleanup Process
Particulates Ash, char, bed material Erosion Filtration, scrubbing
Alkali metals Sodium compounds Corrosion Condensation, ltration,
adsorption
Tars Long-chain aliphatics Deposition on turbine
blades, clogging of
lters
Tar cracking, tar
removal
Fuel nitrogen Ammonia, HCN NO
x
formation Scrubbing, catalytic
reduction
Sulfur, chlorine H
2
S, HCl Corrosion, air pollution Lime scrubbing,
adsorption
Source: Reference 13.
376
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
digester is installed, it is extremely simple to use the biogas it
produces for heating needs such as cooking. Therefore, wide-
spread installation of digesters on small farms could lead to
reduced wood and charcoal use for domestic heat, which would
have benefcial implications Ior indoor air quality, climate
change, quality of life, and deforestation.
Although AD is clearly a mature technology, researchers see
great potential in novel uses for it. In the near term, digester gas
can be upgraded to natural gas quality and injected into regular
natural gas pipelines in places where electricity production is not
appropriate. In the long term, there is great interest in the engi-
neering of microorganisms that can produce hydrogen directly
from anaerobic digestion of biomass.
20,21
This effort is still early
in the research phase, but it shows great promise as an inexpen-
sive source of hydrogen if successful. Another novel concept is
an integrated bioIuelIeedlot Iacility where animal Ieed is a
coproduct of biofuel production and AD is used to augment
power generation. (One such facility is now in commercial oper-
ation in Mead, Nebraska; see www.e3biofuels.com.)
Electricity Generation
Electricity generation technologies, and especially their gas
quality requirements, create some of the key challenges for bio-
energy technology improvements. Such technologies typically
include internal combustion (IC) engines and gas or steam
turbines. (The latter two can be operated together Ior more eIf-
cient combined-cycle generation.) In addition, there have been
several recent installations in which AD gas is used to power
medium-scale molten carbonate fuel cells. Microturbines could
function as possible generators, but this technology is still bat-
tling to overcome durability problems.
The steam turbine is the most common technology currently
used for the generation of power from biomass. Because the
fue gas is not a working fuid, it does not need to meet any
quality requirements when it powers a steam cycle. (Air emis-
sion requirements might apply, however.) Steam cycle eIfcien-
cies are low, between 15% and 25%.
7
In contrast, biomass-based
gas turbines and combined-cycle units require very clean gas
and compression, but they can be up to 60 eIfcient. These
higher eIfciencies explain the great interest in biomass gasif-
cation technologies.
Advantages of IC engines include ready scalability and (up
to 100 kW scale) better tolerance oI high levels oI gas impuri-
ties than gas turbines. However, these parameters vary greatly
depending on engine size and operating temperature. Overall,
engines can be coupled with ambient-pressure gasifers much
more easily than can turbines.
To produce commercially viable large-scale BIGCC plants,
the research agenda must target two key issues: the producer
gas heating value and gas quality requirements for a gas turbine.
The heating value oI the producer gas Irom air-blown gasifers
is insuIfcient to power most contemporary gas turbines. (The
cleaning of producer gas has already been discussed.) Turbine
blades are very sensitive to acids, alkalis, tars, and particulates
entrained in the fuel gas. Reliable data on these tolerances are
diIfcult to obtain, but it is unlikely that turbines can be built to
accept more of these impurities. Hence, the onus is on superior
gas cleaning technology to enable gas turbines to run reliably
on gasifed biomass.
Biofuels
Modern biofuels have been used in some countries since the
late 1970s, as one response to the oil price shocks of that time.
Over 90% of current biofuel production today is ethanol pro-
duced through fermentation, which is described in the sidebar
by Wyman on cellulosic ethanol. Current bioIuels are essen-
tially traditional agricultural commodities that have been put to
a new purpose as fuel, but they are not particularly good fuels.
22

For instance, ethanol has a low energy density, can make fuel
blends corrosive, can increase air pollution in some fuel blends,
and is relatively expensive to produce. (The costs of production
of cellulosic ethanol are unclear because no commercial-scale
plants exist, although six pilot plants are now being built with
the aid oI grants Irom the U.S. Department oI Energy. The expe-
rience gained in their construction and operation will probably
lower production costs signifcantly.) These negative attributes
suggest why bioIuels have needed signifcant government
support in the form of consumption mandates and various sub-
sidies to attain even the relatively minor levels of market
penetration that they have today. Support Ior bioIuels is essen-
tially an agricultural policy designed to provide added income
to producers and processors.
22
Subsidies Ior bioIuels are
estimated to have reached $15 billion in 2007 for the
Organization for Economic Cooperation and Development
(OECD) as a whole.
23
However, the recent rise in the price of crude oil has made
some biofuels economically attractive, which, when combined
with signifcant production subsidies and consumption man-
dates, has led to a rapid expansion of world biofuel use. Current
consumption of biofuels globally constitutes about 2% of trans-
portation fuels by energy content, and this could expand to
about 10% by 2020.
More recently, achieving reductions in GHG emissions has
been added to the rationales for biofuel policies.
24,25
However,
the GHG emissions of current biofuels range from somewhat
lower to greater than those from fossil fuels.
26
Without appro-
priate incentives and regulations that lead to technological
innovation, biofuels are likely to worsen climate change.
2,23
Figure 1 illustrates the fact that dozens of biofuel pathways
are possible and that many of them produce animal feed or elec-
tricity as coproducts. Todays dominant biofuel production
pathways are highlighted in Figure 1: corn ethanol, sugarcane
ethanol, soy biodiesel, and palm biodiesel.
The frst research challenge Ior bioIuels is to address the
problems these compounds entail when used as fuels (as previ-
ously described). It is likely that new compounds with superior
properties and better production processes will emerge.
12,27

Also, note that multiple pathways to biofuel-based hydrocar-
bons, electricity, and hydrogen exist. Because hydrocarbons are
excellent liquid fuels, they might be promising in the short run,
whereas in the long run, electricity and hydrogen pathways
might offer a preferred technology.
The second research challenge for biofuels is related to
their environmental performance. Current biofuel production
can have high GHG emissions, as noted above, and also causes
considerable pollution in some cases, consumes signifcant
quantities of water, and may involve large monocrops of genet-
ically modifed organisms. All oI these issues need to be
addressed.
Most importantly, the currently dominant biofuel produc-
tion pathways all require fertile land because they originate
with agricultural crops, which is unsurprising given how bio-
fuel production is supported. Thus, biofuel feedstock produc-
tion competes with other land uses, such as food production and
wilderness protection. The availability of fertile land is likely
to be a limiting factor (along with water) for the volume of bio-
Iuels that can be produced. (See the sidebar by Gust et al. Ior a
discussion of this issue.) Increasing biofuel production will
induce land use change, including conversion of natural ecosys-
tems into agricultural land (e.g., deforestation). This releases
tremendous amounts of GHGs and threatens ecosystems.
Although the effects of land use change are indirect and often
377
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
ignored, it is increasingly being recognized that, because of the
effects of changes in land use, agricultural-based biofuels have
much higher GHG emissions than is commonly believed.
25,28
Sustainable bioIuels that reduce GHG emissions and protect
ecosystems are likely to require nonagricultural feedstocks,
such as wastes and residues, as well as plants grown on low-
quality land.
29
Another approach might be the increased inte-
gration of biomass feedstock production with food production,
possibly improving nutrient and water management as well.
30
The overall research challenge for biofuels, therefore, is to
develop the technologies for producing more useful compounds
at low cost while minimizing the use of scarce resources such
as arable land and water. This means less reliance on the current
suite of agricultural crops and greater use of wastes, residues,
and feedstocks that can grow on low-quality land. Crucially,
these feedstocks include lignocellulosic materials and algae.
Therefore, the development of low-cost methods for pro-
ducing biofuels from lignocellulosic feedstocks or algae is the
Figure 1. Selected biofuel production pathways with currently dominant pathways highlighted. (Based on a gure drawn by A.D. Jones for
Reference 24.)
Feedstocks Processing Co-Products Fuels
Starches and Sugars Fermentation Ethanol
Sugarcane
Corn Butanol
Sugar Beet
Sweet Sorghum Saccharification
Animal Feed
Ligno-Cellulosic Crops
Switch grass
Miscanthus
High Diversity Grasses
Poplar Gasification Electricity
Ligno-Cellulosic Residues
Corn Stover
Rice Straw Catalysis
Forest Residue and Thinnings Methanol
Bagasse
Orchard Prunings DiMethyl Ether
Ligno-Cellulosic Wastes
Food Waste
Hydrocarbons
Yard Waste
Paper Waste
Flash Pyrolysis
Other Municipal Solid Waste
Construction Debris
Forestry Industry Waste
Oils
Soy
Trans-esterification FAME Biodiesel
Canola (Rapeseed)
Sunflower
Renewable
Palm
Hydrogenation Diesel
Jatropha
Waste Oils
Algae Hydrogen
378
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
key to solving the challenge oI sustainable bioIuels. Such
advances could permit large-scale production of biofuels with-
out using fertile land, there by releasing it for food production
or allowing it to remain as wilderness.
Lignocellulosic production pathways for biofuels include
either biological (fermentation) or thermochemical (e.g., gas-
ifcation) steps. Fermentation by yeasts or other microorgan-
isms requires simple sugars (e.g., glucose), which can be
obtained directly from some feedstocks (e.g., sugarcane) or
through the depolymerization of starch or cellulose. Because
less processing is needed for sugar-bearing plants, biofuels
made from these feedstocks are inherently less costly to pro-
cess (and have lower GHG emissions), followed by starches
and then by lignocellulosic material. Thus, cellulosic feed-
stocks are abundant and cheap, but currently expensive to pro-
cess. This is why many of the key biofuel research issues are
in the area of biotechnology to develop new and improved
methods for overcoming the recalcitrance of cellulose and
decomposing it into constituent sugars (called depolymeriza-
tion). The sidebar by Wyman discusses some key issues.
In addition to biofuels, interest in developing new bioprod-
ucts has also grown recently.
31,32
Although bioproducts might
substitute for products currently based on fossil fuels, there is
no room for a discussion of them here.
Pyrolysis
Pyrolysis is the chemical decomposition of organic materi-
als by heating in the absence of oxygen; it is the process used
to produce charcoal Irom wood and an early step in gasifca-
tion.
33
Biomass pyrolysis products include volatile liquids,
combustible gases, and solid char, the proportions of which are
controlled through the temperature and the rate of reaction. The
combustible gases can be used to generate power or heat, and
the liquids can be converted to biofuel. Recent interest has
focused on the synthesis of biofuels from pyrolysis. In the
1980s, fast pyrolysis was developed as a means of increasing
to 80% the liquid fraction, called bio-oil or bio-crude, which
can be processed further into biofuels. Bio-oil mixtures can
contain more than 400 different compounds, including alcohols,
aldehydes, esters, and aromatic compounds. Fast pyrolysis of
biomass typically occurs at 400900C and takes less than 2 s.
12

During the past two decades, several different reactor designs
have been explored that meet the rapid heat-transfer properties
required.
33
However, there is a great need for the development
of better process controls to consistently achieve target liquid
yields with the desired chemical composition.
Several commercial biomass pyrolysis technologies that
produce bio-oils currently exist. However, bio-oils have signif-
cant drawbacks, including low volatility, high viscosity, coking,
and corrosiveness. Bio-oils must be upgraded or blended to be
used as a transportation fuel. Techniques for upgrading include
hydrogenation, zeolite upgrading, and steam reforming, all of
which use catalysts but none of which have been perfected.
12

Few studies have focused on the use of catalysts for biomass
cracking (in situ upgrading) to generate fuels.
33
Catalysis
A number of different catalytic processes might be impor-
tant in Iuture bioIuel production pathways, including Fischer
Tropsch synthesis (FTS), hydrodeoxygenation oI bio-oils,
steam reIorming, catalytic gasifcation, and biocatalysis.
FTS uses Co-, Fe-, or Ru-based catalysts to produce linear
hydrocarbons and oxygenates, including unrefned gasoline,
diesel, and waxes. This technology was used by Germany dur-
ing the 1930s and 1940s to produce liquid fuels from coal and
is still in use in South AIrica.
34
The product distribution obtained
Irom FTS depends on the catalyst and the process parameters
such as temperature, pressure, and residence time.
34
When
either iron- or cobalt-based catalysts are used in the temperature
range oI 210250C, about 60 oI the FTS product is heavier
than diesel, but this product can be hydrocracked selectively to
yield about 80% aromatic free diesel.
35
The diesel yield can be
increased Iurther by additional processing. Because FTS prod-
ucts are predominantly linear hydrocarbons, they have an
excellent cetane number (about 75 versus the typically required
minimum oI 45). High-temperature fuidized-bed FT reactors
with iron catalysts are ideal Ior the production oI linear olefns,
which are higher-priced petrochemicals.
Biomass gasifcation yields mainly H
2
, CO, CO
2
, and CH
4
,
which is not suitable Ior direct use in FTS but can be tailored
by CH
4
reIorming, watergas shiIt, and CO
2
removal.
UnIortunately, the FTS process appears always to yield prod-
ucts that range all the way from methane (normally unwanted)
to high-carbon-number (up to 22) hydrocarbons.
35
In addition
to olefns and paraIfns, oxygenated products such as alcohols,
aldehydes, and carboxylic acids are also produced. As these
oxygenated products are predominantly linear, it could be of
commercial interest to develop suitable catalysts and operating
conditions to maximize their selectivity.
Hydrodeoxygenation of bio-oils involves treating bio-oils at
moderate temperatures (300600C) with high-pressure H
2
in
the presence of heterogeneous catalysts to remove the oxygen
and Iorm saturated CC bonds. The energy content oI the Iuel
is signifcantly increased, and the stability oI the Iuel increases
during hydrodeoxygenation.
12
Most hydrodeoxygenation work has Iocused on sulfded
CoMo- and NiMo-based catalysts, which are industrial
hydrotreating catalysts for the removal of sulfur, nitrogen, and
oxygen from petrochemical feedstocks. However, bio-oils can
also be upgraded using zeolite catalysts to reduce the oxygen
content and improve the thermal stability. Temperatures of
350500C and atmospheric pressure are used Ior zeolite
upgrading. The advantages of using a zeolite catalyst are that
no H
2
is required, atmospheric processing reduces operating
cost, and the temperatures are similar to those for bio-oil pro-
duction. However, poor hydrocarbon yields and high yields of
coke limit the usefulness of zeolite upgrading.
12
Steam reIorming oI Iossil Iuels is a well-established tech-
nology, and steam reforming of bio-oils is an extension of this
technology. Steam reIorming reactions are perIormed at high
temperature (600800C) and short residence time, usually
with a Ni-based catalyst. This reaction predominantly produces
CO and H
2
syngas. Commercial catalysts consist essentially of
Ni supported on alumina. Steam reIorming oI bio-oils is com-
plicated because some bio-oil components are thermally unsta-
ble and decompose upon heating. Deactivation of the catalysts
due to coking is one of the major problems, and bio-oils have
more deactivation problems than do petroleum-derived oils. To
address some of the above issues, a high temperature is needed
in the reactor to gasify coke deposits formed by thermal decom-
position, and large quantities of steam are necessary to avoid
catalyst deactivation by coking.
12
In addition to traditional noncatalytic gasifcation processes
(described above), catalytic gasifcation oI biomass can be used
to lower the operating temperature, making the process more
economically feasible. Typically, temperatures above roughly
750C are required Ior noncatalytic gasifcation. Gasifcation
catalysts can lower the operating temperature to around 600C,
resulting in substantial energy savings. A common gasifcation
catalyst is dolomite [CaMg(CO
3
)
2
], an inexpensive carbonate.
However, rapid deactivation by tar formation at the lower tem-
peratures oI catalytic gasifcation has led to the search Ior more
379
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
tar-resistant, but still active, gasifcation catalysts. Ni-based
catalysts have been widely studied, but they are usually based
on steam reforming formulations and have not been optimized
Ior catalytic gasifcation reactions and still suIIer Irom deacti-
vation by tars.
36
Catalysis research can also aid in more eIfcient gas cleanup
Irom gasifcation. Currently, catalysts Ior the decomposition oI
tars downstream oI the gasifer (as opposed to in-bed tar
removal) are typically based on Ni catalysts, which operate at
700900C. These catalysts are subject to deactivation, but rel-
atively few systematic studies of the deactivation process or
how to regenerate the catalyst, particularly in situ, have been
performed. Improvements in these technologies could be
important.
36
One other objective of catalysis research in gas
cleanup is to intensify the processes/materials required to
remove all of the contaminants, simplifying them into as few
continuous operations as possible. Examples include combin-
ing alkali and particulate removal and catalytic tar and ammo-
nia decomposition on a particulate-resistant support.
36
Specifc research areas Ior catalysis include the Iollowing:
N FT catalysts and/or process conditions with better selectivity
for carbon number. Meeting these objectives would require
the successful manipulation of the FT chain-growth mecha-
nism. Unfortunately, to date, none of the hotly contested pro-
posed FT mechanisms have indicated how this goal could be
achieved.
35
N Cost-eIIective gasifcation catalysts that are resistant to deac-
tivation and that will help improve the overall economics.
36
N Novel catalytic reactor designs tailored to the typically
smaller scale of biomass conversion processes.
36
N Mechanically stronger FT catalysts to reduce erosion.
N Catalysts for downstream adjustment of the H
2
/CO ratio for
specifc end products and Ior the reIorming oI biomass-
derived liquids (e.g., from pyrolysis).
N Production of hydrogen-rich gases for subsequent hydrogen-
powered energy devices.
Novel Concepts
Several novel concepts are now emerging in the area oI bio-
Iuels. One is the potential oI using Iree-foating algae to pro-
duce biofuels, perhaps similarly to how they are used today to
produce food supplements such as spirulina.
3740
Most research
efforts in this area are biological in nature, but research needs
include materials and processes to maintain adequate growing
conditions and catalysts or other processes to convert algae-
derived materials into suitable fuels.
Microbial fuel cells (MFCs) are a special class of fuel
cells in which biocatalysts such as microorganisms or enzymes
are employed instead of metallic inorganic catalysts.
41,42
Such
devices were discovered over a century ago and largely aban-
doned, but recent advances in nanoscale science and technol-
ogy might make it possible to overcome key drawbacks, such
as short lifetime and low power density, that have limited
enzyme-based biofuel cells from being used for practical
applications. Various nanostructures exhibit the potential to
stabilize and activate enzymes with much improved perfor-
mance. The large surface areas that nanostructures provide
for the attachment of enzymes can increase the enzyme load-
ing and possibly improve the power density of biofuel
cells. In that sense, nanoscale engineering of the biocatalysts
appears to be critical in the next stage of advancement of bio-
fuel cells.
One innovative approach would be to join biomass and coal
gasifcation with carbon capture and sequestration (CCS) in a
single production pathway.
43
Several studies have shown that
the generation of negative atmospheric carbon emissions by
integrating CCS and biomass energy technologies is potentially
economical. The ability oI biomassCCS to generate negative
atmospheric carbon emissions could fundamentally change the
role of biomass in achieving deep emissions reductions by pro-
viding a mechanism to offset emissions anywhere in the econ-
omy.
44,45
Recently, there has been extensive discussion of the
possibility oI using a biomass integrated gasifcation Fischer
Tropsch (BIGFT) process.
46
A BIGFT plant would cogene-
rate electricity, heat, and fuel from biomass and potentially
lower GHG emissions signifcantly.
Another novel concept is the use of nuclear process heat in
biofuel production, leaving more of the scarce biomass to be
turned into fuel.
47
Further, if nuclear-derived hydrogen were
available, all of the carbon could be converted to high-quality
hydrocarbon liquid fuels, possibly doubling the volume of
available biofuels.
Bioenergy Supply Projections
The predicted volume of bioenergy in the global supply over
the next half-century is much contested.
2,48
The wide variation
in projections is not surprising given the wide range of forces
that can drive bioenergy use. Key uncertainties include differ-
ing predictions regarding the state of bioenergy and fossil fuel
technologies, feedstock costs, the rate of global economic
growth, climate policy, and the need for and availability of land
for energy crops, among others.
Berndes et al.
48
usefully categorized studies of bioenergy
supply as shown in Table II. Demand-driven studies assess the
demand for bioenergy that is determined by policy, economic,
and environmental drivers. Resource-focused studies estimate
the bioenergy resource base that can be feasibly extracted.
Table II shows the upper and lower bounds of both types of
studies, which suggest that the share of bioenergy might pro-
vide 540 oI the total primary energy supply in 2050.
However, resource-focused assessments assume that energy
crops make up the majority contribution to the bioenergy sup-
ply. The potentials for wastes, biomass production on degraded
lands, and novel concepts such as algae-derived biofuels have
not been considered in these studies, although they could be
substantial. For instance, a recent evaluation found that, in the
United States, municipal solid waste and Iorestry residues could
contribute the equivalent oI 1030 oI annual gasoline
demand.
49
Thus, it appears that bioenergy can eventually supply a sub-
stantial fraction of world energy (up to several tens of percent).
However, for such levels to be environmentally sustainable and
to not adversely affect the food security of the poor, improve-
ments in both technologies and policies are needed.
2
Conclusions
Biomass remains a key energy supply for several billion
people living in developing countries, and with near-record oil
prices, the production of liquid biofuels for transportation is
growing rapidly. However, both traditional biomass energy and
Table II: Bioenergy Supply Projections in 2050.
Demand-Driven
Studies
Resource-Focused
Studies
Low estimate (EJ/
year)
22 47
High estimate (EJ/
year)
320 460
Note: Global energy demand is about 470 EJ/year.
380
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
crop-based biofuels technologies have negative environmental
and social impacts that are unlikely to continue to be acceptable.
Therefore, technological innovation will be needed in bioen-
ergy, and changes in public policies will be equally critical.
Most immediately, enhancements in cookstoves can have
multiple benefts such as improved indoor air quality and
improved quality of life for the rural poor in developing coun-
tries. Further, resources that are diffuse or are currently wasted,
such as grasses and organic waste, could, with improved tech-
nology, serve as fuels for distributed electricity generators that
supply the rural poor.
In terms of biofuels technologies, the most important areas
for research are those that enable the use of lignocellulosic
feedstocks and yield high-quality liquid fuels, such as bio-based
hydrocarbons. The ability to use lignocellulosic feedstocks, in
turn, permits biofuel production without the use of productive
land, releasing the land for food production or allowing it to
remain as wilderness. Other biofuel technologies that require
little or no arable land, such as algae, are similarly important.
If carbon capture and sequestration technologies can be
successfully applied, bioenergy production could become
carbon-negative, creating a potentially lucrative opportunity.
Fortunately, many possible production pathways from lignocel-
lulosic feedstocks to high-quality biofuels could become
feasible with materials (and other) research.
Technological progress will not be suIfcient, however, as
agricultural land might be used to grow energy crops instead of
food crops because of the higher market value of the former,
and the implications such land use can have for food prices
appears to be adverse.
50,51
This could force the poorest commu-
nities to compete for food with the richest attempting to buy
fuel. Moreover, under current policies, the environmental
effects from biofuel production might be worse than those from
fossil fuels.
With improved policies and technologies, however, bioen-
ergy could contribute signifcantly to economic, social, and
environmental goals. Eliminating the traditional mindset of
biofuels as simply a new market for agricultural crops, subject
to major support policies such as subsidies, will be essential to
the development of cost-competitive biofuels. Policies that
limit environmental impacts of biofuels, such as deforestation
and greenhouse gas emissions, will help advance improve-
ments in technology. The increased demand for biomass for
energy could potentially improve farm incomes, and with well-
designed policies, it could beneft the poorest Iarmers.
References
1. A. Demirbas, Energy Sources A: Recovery, Utilization, Environ. Effects 29,
303 (2007).
2. A.D. Sagar, S. Kartha, Annu. Rev. Energy Environ. 32, 131 (2007).
3. J.P. McCracken, K.R. Smith, Environ. Int. 24, 739 (October 1998).
4. A.V. Bridgewater, Fuel 74, 631 (1995).
5. D.A. Tilman, Biomass Bioenergy 19, 365 (2000).
6. A.L. Robinson, J.S. Rhodes, D.W. Keith, Environ. Sci. Technol. 37, 5081
(November 15, 2003).
7. P. Quaak, H. Knoef, H. Stassen, Energy from Biomass: A Review of
Combustion and Gasication Technologies (The World Bank, Washington,
DC, 1999).
8. Multiple Links between Household Energy and the Millenium Development
Goals (World Health Organization, Geneva, Switzerland, 2007).
9. M. Ezzati, D.M. Kammen, Annu. Rev. Energy Environ. 27, 233 (2002).
10. L. Baxter, F0e| 84, 1295 (2005).
11. A. Faaij, R. van Ree, L. Waldheim, E. Olsson, A. Oudhuis, A. van Wijk, C.
Daey-Ouwens, W. Turkenburg, Biomass Bioenergy 12, 387 (1997).
12. G.W. Huber, S. Iborra, A. Corma, Chem. Rev. 106, 4044 (September
2006).
13. A.V. Bridgewater, Fuel 74, 631 (1995).
14. C. Higman, M. van der Burgt, Gasication 391 (Elsevier, San Francisco,
CA, 2003).
15. A.C. Caputo, M. Palumbo, P.M. Pelagagge, F. Scacchia, Biomass Bioenergy
28, 35 (2005).
16. D.A. Burke, Dairy Waste Anaerobic Digestion Handbook 57 (Environmental
Energy Company, Olympia, WA, 2001); www.makingenergy.com.
17. P. Borjesson, M. Berglund, Biomass Bioenergy 30, 469 (2006).
18. P. Borjesson, M. Berglund, Biomass Bioenergy 31, 326 (2007).
19. E. Ghafoori, P.C. Flynn, M.D. Checkel, Int. J. Green Energy 4, 339 (2007).
20. Z. Bagi, N. cs, B. Blint, L. Horvth, K. Dob, K. Perei, G. Rkhely, K.
Kovcs, Appl. Microbiol. Biotechnol. 76, 473 (2007).
21. M. Cooney, N. Maynard, C. Cannizzaro, J. Benemann, Bioresour. Technol.
98, 2641 (2007).
22. D.M. Kammen, A.E. Farrell, R.J. Plevin, A.D. Jones, M.A. Delucchi, in
OECD Research Round Table: Biofuels: Linking Support to Performance, S.
Perkins, Ed. (Organisation for Economic Cooperation and Development,
Paris, 2007), p. 28.
23. Joint Transport Research Center, Biofuels: Linking Support to Performance
(Organisation for Economic Cooperation and Development, Paris, 2007);
www.cemt.org/JTRC/EconomicResearch/RoundTables/RTbiofuelsSummary.
pdf.
24. A. Schwarzenegger, Executive Order S-01-07: Low Carbon Fuel Standard
(Sacramento, CA, 2007).
25. A.E. Farrell, D. Sperling, A.R. Brandt, B.K. Haya, J. Hughes, B.M. Jenkins, A.D.
Jones, D.M. Kammen, C.R. Knittel, M.W. Melaina, J.M. Ogden, M. OHare, R.J. Plevin,
A Low-Carbon Fuel Standard for California Part 1: Technical Analysis (Transportation
Sustainability Research Center, University of California, Berkeley, CA, 2007); http://
repositories.cdlib.org/its/tsrc/UCB-ITS-TSRC-RR-2007-2.
26. A.E. Farrell, R.J. Plevin, B.T. Turner, A.D. Jones, M. OHare, D.M. Kammen,
Science 311, 506 (January 27, 2006).
27. C.A.C. Sequeira, P.S.D. Brito, A.F. Mota, J.L. Carvalho, L. Rodrigues, D.
M.F. Santos, D.B. Barrio, D.M. Justo, Energy Convers. Manage. 48, 2203
(July 2007).
28. A.E. Farrell, D. Sperling, A.R. Brandt, A. Eggert, B.K. Haya, J. Hughes, B.M.
Jenkins, A.D. Jones, D.M. Kammen, C.R. Knittel, M.W. Melaina, M. OHare, R.J.
Plevin, A Low-Carbon Fuel Standard for California Part 2: Policy Analysis
(Transportation Sustainability Research Center, University of California, Berkeley,
CA, 2007); http://repositories.cdlib.org/its/tsrc/UCB-ITS-TSRC-RR-2007-3/.
29. D.A. Tilman, J. Hill, C. Lehman, Science 314, 1598 (December 8, 2006).
30. R.P. Anex, L.R. Lynd, M.S. Laser, A.H. Heggenstaller, M. Liebman, Crop
Sci. 47, 1327 (JulyAugust 2007).
31. R.D. Perlack, L.L. Wright, A. Turhollow, R. Graham, B. Stokes, D. Erbach,
Biomass as a Feedstock for a Bioenergy and Bioproducts Industry: The
Technical Feasibility of a Billion-Ton Annual Supply (Oak Ridge National
Laboratory, Oak Ridge, TN, 2005).
32. W.E. Mabee, D.J. Gregg, J.N. Saddler, Appl. Biochem. Biotechnol. 121,
765 (Spring 2005).
33. D. Mohan, C.U. Pittman, P.H. Steele, Energy Fuels 20, 848 (May 2006).
34. A. Demirbas, Prog. Energy Combust. Sci. 33, 1 (February 2007).
35. M.E. Dry, Appl. Catal. A 276, 1 (November 2004).
36. J.J. Spivey, Catal. Today 100, 171 (February 2005).
37. A.P. Carvalho, L.A. Meireles, F.X. Malcata, Biotechnol. Prog. 22, 1490
(December 2006).
38. J.M. Gordon, J.E.W. Polle, Appl. Microbiol. Biotechnol. 76, 969 (October
2007).
39. B. Hankamer, F. Lehr, J. Rupprecht, J.H. Mussgnug, C. Posten, O. Kruse,
Physiol. Plant. 131, 10 (September 2007).
40. N. Ladygina, E.G. Dedyukhina, M.B. Vainshtein, Process Biochem. 41,
1001 (May 2006).
41. Z.W. Du, H.R. Li, T.Y. Gu, Biotechnol. Adv. 25, 464 (SeptemberOctober
2007).
42. J. Kim, H.F. Jia, P. Wang, Biotechnol. Adv. 24, 296 (MayJune 2006).
43. R.H. Williams, E.D. Larson, H. Jin, 8rh International Conference on
Greenhouse Gas Control Technologies, 6 (Trondheim, Norway, 2006).
44. C. Azar, K. Lindgren, E. Larson, K. Mollersten, Clim. Change 74, 47
(January 2006).
45. J.S. Rhodes, D.W. Keith, Biomass Bioenergy 29, 440 (2005).
46. M.J.A. Tijmensen, A.P.C. Faaij, C.N. Hamelinck, M.R.M. van Hardeveld,
Biomass Bioenergy 23, 129 (2002).
47. C.W. Forsberg, Assessment of NuclearHydrogen Synergies with
Renewable Energy Systems and Coal Liquefaction (Oak Ridge National
Laboratory, Oak Ridge, TN, 2006); p. 41.
48. G. Berndes, M. Hoogwijk, R. van den Broek, Biomass Bioenergy 25, 1
(2003).
49. A.D. Jones, M. OHare, A.E. Farrell, Biofuel Boundaries: Estimating
the Medium-Term Supply Potential of Domestic Biofuels (Transportation
Sustainability Research Center, University of California, Berkeley, CA, 2007);
http://repositories.cdlib.org/its/tsrc/UCB-ITS-TSRC-RR-2007-4.
50. D.J.A. Johansson, C. Azar, Clim. Change 82, 267 (June 2007).
51. U.A. Schneider, B.A. McCarl, Environ. Res. Econ. 24, 291 (April 2003). N
381
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
What Is Cellulosic Biomass?
Although ethanol is now made from the sugars in the starch
fraction of corn and other crops and from the sugar in sugar-
cane, a much greater impact for ethanol in terms of fuel use
could be realized if the sugars from more recalcitrant cellulosic
biomass could be converted to ethanol. Cellulosic biomass is
the structural portion of plants and includes agricultural (e.g.,
corn stover, which is all of the above-ground portion of the
corn plant, excluding the grain) and forestry (e.g., sawdust)
residues, major fractions of municipal solid waste (e.g., waste
paper and yard waste), and herbaceous (e.g., switchgrass) and
woody (e.g., poplar) crops grown as energy resources.
1

Although distinctive in outward appearance, these materials all
comprise about 4050 cellulose and 2030 hemicellulose,
with lesser amounts of lignin and other compounds such as
sugars, oils, and minerals. Cellulose is a polymer of glucose
sugar molecules that are physically linked together in a crystal-
line structure to provide structural support for plants.
Hemicellulose is also made up of sugars covalently joined
together in long chains, but it generally includes fve diIIerent
sugars: arabinose, galactose, glucose, mannose, and xylose. In
addition, hemicellulose is an amorphous, branched material.
Lignin is a phenylpropene compound that can be viewed as a
low-sulfur, immature coal.
Biological Conversion of Cellulosic Biomass to
Ethanol
The overall approach to converting cellulosic biomass
to ethanol is outlined in Figure 1. Acids can break down
the long chains in hemicellulose and cellulose to release
the sugars comprising these materials through hydrolysis
reactions, but because of their high specificity, enzymes
known as cellulase can achieve higher yields of glucose
from cellulose and are often favored.
2
However, to survive
in nature, cellulosic biomass has evolved a structure that
resists enzymatic attack, and it must first be pretreated to
better expose the cellulose to enzymes. Over the years, bio-
logical, chemical, mechanical, physical, and thermal pre-
treatments have been tried, but currently, only methods that
treat biomass at temperatures oI about 80200C with one
or more chemicals have been able to realize the high sugar
yields vital to economic success. Leading thermochemical
pretreatment options include dilute sulfuric acid, ammonia
fiber expansion (AFEX), neutral pH, and lime, although
the choice depends on many factors including the nature of
the substrate, enzymes, and organisms and other process
details.
3,4
A portion of pretreated biomass can be used to feed a fun-
gus or other organism that produces cellulase that can then be
added to the bulk of the pretreated solids to release glucose
from cellulose. In addition, the enzymes must be capable of
releasing sugars contained in hemicellulose if the pretreat-
ment step does not fully accomplish this. Then, an organism
is added to ferment all of the sugars to ethanol. Although con-
ventional yeasts ferment glucose and other sugars containing
six carbon atoms eIfciently, they do not use the fve-carbon
sugars arabinose and xylose well, and bacteria and yeasts have
been genetically modifed to accomplish this task. Following
fermentation, the broth then passes into a distillation system
for recovery and concentration of the ethanol, while the lignin
and other portions not converted into ethanol can be burned to
generate all of the heat and electricity needed to run the entire
process, with excess power left to export to homes and busi-
nesses. The sugars released during pretreatment and/or enzy-
matic hydrolysis of hemicellulose and cellulose could also be
biologically or chemically converted into other products in
addition to ethanol, and lignin, an aromatic compound, could
be employed to produce various materials such as adhesives
and fllers.
5
Benets of Cellulosic Ethanol
Petroleum provides more energy internationally than any
other source, accounting for about 35% of total energy pro-
duced, and unstable regions of the world hold the bulk of all
known reserves. In addition, because over half of the petro-
leum is used for a transportation sector that is almost totally
dependent on petroleum, developing alternative sources of
liquid transportation fuels is vital to alleviating this perilous
dependence,
6
with biomass uniquely suited to sustainably
meet this need. As one measure oI its abundance, the U.S.
Department oI Agriculture (USDA) and Department oI Energy
(DOE) recently projected that about 1.3 billion dry tons of
cellulosic biomass could be produced annually in the United
States, Irom which enough ethanol could potentially be made
to replace over half of the gasoline now used in the country.
7

Cellulosic Ethanol: A Unique Sustainable Liquid
Transportation Fuel
Charles E. Wyman (University of CaliforniaRiverside, USA)
Figure 1. Overall process ow diagram illustrating the important
operations in the biological conversion of cellulosic biomass to ethanol.
Biological steps to
produce enzymes,
hydrolyze biomass,
and ferment sugars
Cellulosic
biomass
Pretreatment to
open up biomass
to enzymes
Ethanol recovery
and purification
Residue processing
Boiler and
power generation
Fuel ethanol
Process effluents
Exported
electricity
Lignin and
other residues
Process heat
and power
382
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
Thus, cellulosic ethanol production could play a major role in
signifcantly reducing petroleum imports, improving energy
security, creating rural agricultural and manufacturing jobs,
and dramatically cutting the trade defcit Ior oil imports.
Ethanol has higher octane content than regular or even pre-
mium gasoline; this property, coupled with a higher heat of
vaporization, permits operation with a higher compression
ratio when ethanol is used as a pure fuel. The result can be
considerably better eIfciency compared to gasoline in terms
of miles powered per unit energy content of the fuel, compen-
sating to some degree for the somewhat lower energy content
of ethanol compared to gasoline. Thus, ethanol must sell for
about 6780 oI the price oI gasoline to provide equivalent
cost per distance traveled. However, a larger portion of etha-
nol is used as blends with gasoline, and all automobile manu-
facturers warrant their vehicles for this use and employ
appropriate materials in fuel lines to accommodate the some-
what greater corrosivity of blended ethanol. Although pure
ethanol has low evaporative losses that can cause smog, low-
level blends with gasoline increase evaporation somewhat,
potentially negating some oI the benefts Ior reducing tailpipe
emissions.
In addition to the powerful strategic and economic attributes
of cellulosic ethanol, the use of lignin to power the cellulosic
ethanol production facility and the low fossil fuel inputs
required to plant, grow, harvest, and transport cellulosic bio-
mass and to transport ethanol to its destination result in very
favorable energy balances for cellulosic ethanol. As a result,
less than 10% of the ethanol energy produced has required fos-
sil energy inputs.
8
Even more importantly, these low fossil fuel
requirements translate into very low net greenhouse gas emis-
sions of less than 12% of those of gasoline;
8
this ratio could be
improved if the electricity exported to the power grid displaced
requirements for coal and more renewable fuels were employed
in all operations.
9
In addition, more carbon dioxide could be
removed than released in the process if the virtually pure carbon
dioxide produced during fermentation could be sequestered
in underground caverns or through other techniques being
developed.
10
Technical and Economic Progress
Cellulosic biomass is a low-cost feedstock, with a cost of
about $40/dry ton being competitive with petroleum at about
$13/barrel on an equivalent-energy-content basis.
11
Although
the resistance of biomass to biological breakdown has been his-
torically expensive to overcome, an important advantage to
using biological processes for this operation is that we can
employ the power of modern biotechnology to dramatically
improve the process and reduce processing costs.
12
In this vein,
advances in pretreatment, enzymes, and fermentative organ-
isms as well as process integration have lowered the costs from
about $45/gal oI ethanol in the early 1980s to being competi-
tive with ethanol made from corn now.
13
Furthermore, costs as
low as about $0.60/gal could be realized if leap-forward tech-
nology advances were made in the most expensive processing
steps: pretreatment and the biological conversion steps of
enzyme production, enzymatic hydrolysis, and fermenta-
tion.
14,15
For example, genetically modifying organisms to make
their own enzymes while also fermenting the sugars they release
to ethanol with high yields is one promising approach. Another
would be advanced pretreatment systems, or plants that are
genetically modifed to be more susceptible so that just water
could be used for pretreatment, thus avoiding the need for
exotic materials of construction. Advances in low-cost materi-
als of construction to handle dilute acid or other chemicals
would also be invaluable.
Commercial Challenges
Despite its tremendous potential benefts and ongoing
technical and economic progress, cellulosic ethanol technol-
ogy has yet to be commercialized. First, enzymes have been
too expensive historically, but through Iunding Irom the U.S.
DOE, both Novozymes and Genencor claim to now be able to
produce low-cost enzymes.
16,17
In addition, although operating
costs can be low for technology that achieves high ethanol
yields, capital costs are high, and private investors are reluc-
tant to undertake such large projects Ior frst-oI-a-kind opera-
tions. The fact that ethanol is a commodity product with tight
margins further impedes commercialization. Thus, the inher-
ently high capital costs coupled with the high cost of capital
Ior frst projects is the major obstacle.
15
Various opportunities
can be employed to counter this risk, such as retroftting exist-
ing Iacilities, fnancing with low-cost government bonds, sell-
ing higher value coproducts, and using low-cost wastes as
feedstocks. However, although cellulosic ethanol can ulti-
mately compete without subsidies, government assistance and
policy will likely be critical to overcoming the concerns about
risk Ior frst applications and getting the industry oII the
ground, just as it was in facilitating the emergence of the pet-
rochemical industry during World War II. To be eIIective,
incentives Ior the frst Iew commercial Iacilities must be per-
manent so that large capital investments cannot be stranded by
changes in government policy. Improving our knowledge of
the fundamentals of the complex interaction of solid cellu-
losic biomass with enzymes and organisms would also build
confdence in technology applications and suggest routes Ior
dramatically advancing the technology to reduce costs.
Closing Thoughts
The production of ethanol from low-cost and abundant cel-
lulosic biomass provides a powerful and unique route to sus-
tainable production of liquid transportation fuels that our
society uses so extensively and offers tremendous economic,
strategic, and environmental benefts. Signifcant progress has
been made in advancing the technology so that it is cost-
competitive now, and lower costs are foreseeable. However, the
challenge to widespread use and commercialization is over-
coming the risk oI frst applications, with well-directed govern-
ment policies likely being critical to the emergence of this
industry, an industry that can have a major impact on our quest
to reduce use of petroleum from unstable regions of the world
and also cut greenhouse gas emissions that present a mounting
threat to our way of life.
Acknowledgments
The author is grateful to the Ford Motor Company for spon-
soring the Chair in Environmental Engineering at the Bourns
College of Engineering at the University of California,
Riverside, and for providing funds for our research on cellu-
losic ethanol.
References
1. C.E. Wyman, S.R. Decker, M.E. Himmel, J.W. Brady, C.E. Skopec, L. Viikari,
in Polysaccharides: Structural Diversity and Functional Versatility, S. Dumitriu,
Ed. (Dekker, New York, ed. 2, 2004), p. 995.
2. C.E. Wyman, in Encyclopedia of Energy, C. Cleveland, Ed. (Elsevier,
St. Louis, MO, 2004), p. 541.
3. C.E. Wyman, B.E. Dale, R.T. Elander, M. Holtzapple, M.R. Ladisch, Y.Y. Lee,
Bioresour. Technol. 96 (18), 2026 (2005).
383
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
4. N. Mosier, C.E. Wyman, B. Dale, R. Elander, Y.Y. Lee, M. Holtzapple,
M. Ladisch, Bioresour. Technol. 96 (6), 673 (2005).
5. C.E. Wyman, Applied Energy Technology Series, 424 (Taylor and Francis,
Washington, DC, 1996).
6. U.S. Department of Energy, Energy Information Administration, Annual
Energy Review 2005 (DOE Report DOE/EIA-0384, 2005).
7. R. Perlack, L. Wright, A. Turhollow, R. Graham, B. Stokes, D. Erbach,
Biomass as Feedstock for a Bioenergy and Bioproducts Industry: The
Technical Feasibility of a Billion-Ton Annual Supply (Oak Ridge National
Laboratory, Oak Ridge, TN, 2005).
8. A.E. Farrell, R.J. Plevin, B.T. Turner, A.D. Jones, M. OHare, D.M. Kammen,
Science 311, 506 (2006).
9. C.E. Wyman, Appl. Biochem. Biotechnol. 456, 897 (1994).
10. H.S. Kheshgi, R.C. Prince, Energy 30, 1865 (2005).
11. L.R. Lynd, C.E. Wyman, T.U. Gerngross, Biotechnol. Prog. 15 (5), 777
(1999).
12. U.S. Department of Energy, Ofce of Science, Breaking the Biological
Barriers to Cellulosic Ethanol, a Joint Research Agenda, a Research Roadmap
Resulting from Biomass-to-Biofuels Workshop, Rockville, December 2005
(DOE/SC 0095, June 2006).
13. C.E. Wyman, Annu. Rev. Energy Environ. 24, 189 (1999).
14. L.R. Lynd, R.T. Elander, C.E. Wyman, Appl. Biochem. Biotechnol.
5758, 741 (1996).
15. C.E. Wyman, Trends Biotechnol. 25 (4), 153 (2007).
16. Chem. Eng. Prog. 100 (12), 15 (2004).
17. Chem. Eng. News 83 (17), 10 (April 25, 2005). N
All oxygen-dependent life depends on photosynthesis. In
addition to breathing the oxygen produced by photosynthesis,
humans have been harnessing energy from photosynthesis for
millennia. Since the beginning oI human societal structures,
human needs have driven the evolution of agricultural produc-
tion, and they continue to do so. Recently, it has been suggested
that agriculture can contribute substantially to human techno-
logical (nonnutritional) energy needs. This possibility raises
concern because the projections of human energy needs argue
convincingly that without large increases in energy conversion
eIfciency (ECE), land-grown bioIuel production and Iood pro-
duction will compete for land, a largely untenable compromise
given the current nutritional status of the worlds underdevel-
oped societies.
In addition to using the fuel provided by natures photosyn-
thetic process, humans have devised direct routes for harness-
ing solar energy including, for example, photovoltaic (PV)
cells. These cells produce energy in the form of electromotive
force (emf, electricity), which, although ideal for many applica-
tions, is not easily stored and used for fuel (e.g., in transporta-
tion). We posit that transIormational progress toward meeting
the goals of supplanting fossil fuels, providing energy security,
and mitigating climate change can be made at the intersection
oI technology and biology. This intersection comprises artifcial
photosynthesis, other bio-inspired energy conversion processes,
and the design oI organisms that specialize in eIfcient bioIuel
production Irom solar energy. As outlined here, artifcial con-
structs can contribute directly to solar energy conversion, can
be incorporated into hybrid systems, and can inform the design
of new photosynthetic organisms.
What Do We Mean by Ecient, and Why Isnt
Natural Photosynthesis More Ecient?
The initial energy-conserving steps in the conversion of
solar energy to either electricity or biomass can be described by
elementary photophysical processes; the essential ones are
shown in Figure 1. The absorption of light (red and green
arrows) promotes an electron to a higher energy level, which
leads to an excited state in which an electron is repositioned in
spatial and energy coordinates and a positive charge (hole) is
left behind. This is the transformation of solar to chemical
energy; the electron is chemically reducing (low electrochemi-
cal potential), and the hole is chemically oxidizing (high
electrochemical potential). In molecular systems, the further
stabilization necessary to prevent wasteful relaxation back to
the ground state involves moving the electron and hole farther
apart; there is a concomitant loss of energy (illustrated by the
dash-dotted arrows in Figure 1) necessary to drive this charge
separation process. In typical PV cells, the hole and electron are
separated and thereby stabilized by an internal electric feld at
the junction of the n- and p-type semiconductor materials.
The energy associated with separating the charges (dash-dotted
arrows in Figure 1) reduces the electrical energy available
in the external circuit. Charge separation sets the stage for
describing three eIfciency-defning processes: a high Iraction
of the photons absorbed must yield charge separation (i.e., the
quantum yield of charge separation must be high); the energy
of the charge-separated state must be high; and recombination
of the electron and hole, producing heat, must be much slower
than chemical reactions making productive use of the oxidation
and reduction potential (or slower than the conduction of charge
in a PV device).
ECE is defned as the usable electrical or harvestable chemi-
cal energy output divided by the total solar energy incident on the
organism or device. In terms of meeting human energy needs,
which are usually expressed on an annual basis, it is convenient
to calculate ECE using insolation (incident solar energy) per year
summed over diurnal and seasonal cycles. ECE is a fundamental
parameter that determines the area required to provide a specifed
amount oI energy Ior human use. Some examples oI the ECEs oI
Engineered and Articial Photosynthesis: Human
Ingenuity Enters the Game
Devens Gust (Arizona State University, USA), David Kramer (Washington State University, USA),
Ana Moore (Arizona State University, USA), Thomas A. Moore (Arizona State University, USA),
and Wim Vermaas (Arizona State University, USA)
384
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
biofuel production by photosynthetic organisms and electrical
work by a typical PV cell are listed in Table I.
Clearly, not all of the energy of incident sunlight can be
converted into useIul work. In addition to the eIfciency-
limiting processes mentioned above, other factors come into
play. The ECE is limited frst by the Iraction oI sunlight
absorbed by the organism or device: photons having less
energy than required for the lowest-energy absorption transi-
tion of the material will not be absorbed. This is illustrated by
the pink arrow in Figure 1. The ECE is also limited because
both photosynthetic organisms and simple PV devices such as
silicon solar cells function as single-threshold systems. In
other words, the energy of absorbed photons that is above the
lowest excited singlet state of chlorophyll (or the conduction-
band energy of a solar cell) is lost as heat during the electronic
relaxation processes that populate the lowest excited state (see
arrow labeled Internal conversion in Figure 2 and the dashed
arrow in Figure 1). In photosynthesis, this leads to a loss of
about one-fourth of the absorbed energy. A threshold or band-
gap of about 1.3 eV is optimal. The increase in chemical poten-
tial of the charge-separated state with increasing bandgap
comes at the expense of the fraction of photons (in the pink
arrow category in Figure 1) absorbed from the solar spectrum.
The bandgap in water-oxidizing photosynthesis is about 1.8 eV,
well on the high-energy side of optimal. Finally, as mentioned
above, conversion devicesnatural and artifcialsacrifce
some of the potentially available energy in order to slow charge
recombination reactions and
drive desired chemical or elec-
trical processes forward. As a
result of these factors, illustrated
by the dash-dotted arrows in
Figure 1, a signifcant Iraction
of photon energy must be lost as
heat in any conversion device.
The best single-threshold photo-
voltaic devices have ECEs that
approach the ShockleyQuiesser
limit of about 30%, which was
calculated taking all of the above
considerations into account. In
contrast, although the initial
steps of plant and bacterial
photosynthesis often have very
high quantum yields, the ECE of
natural photosynthesis is rela-
tively low (maximally about
6% but usually observed to be
<0.8%).
The apparent paradox of a
high quantum yield and low ECE
in natural photosynthetic systems
can be resolved by analyzing the
energy fow in the photosynthetic
process presented in Figure 2,
which qualitatively illustrates the energy losses from absorbed
incident radiation during photosynthesis by a typical terrestrial
plant leaf at full sunlight. Photosynthetic organisms have evolved
by maximizing fexibility in metabolism and nutrient supply and
are therefore able to adapt to diverse environmental conditions
that are not necessarily consistent with overall energy storage
eIfciency. The physiological reactions required Ior steady-state
maintenance of energized membranes and reproduction (that is,
the biosynthesis of components that are not particularly energy-
rich and the turnover of cellular components) are essential to the
survival of the organism (Figure 2, arrow labeled Growth and
maintenance). These processes require copious amounts of
energy and resources, but do not directly produce harvestable,
high-energy-content biomass for human use.
The evolution of natural photosynthesis was not driven by
maximally eIfcient energy storage, but rather allowed organ-
isms to tap into a new power source, solar energy, and thus to
colonize and survive in new environments far from the geo-
chemical energy sources that likely powered early life. In this
view, natural photosynthesis can be thought of as an add-on
module that probably appeared after many of the central meta-
bolic and bioenergetic processes had been established. Although
the ferce evolutionary struggle oI photosynthetic organisms to
prevail in their local surroundings has certainly tuned photo-
Lowest excited state
Ground state
Bandgap
or
threshold
E
n
e
r
g
y
Work out
Figure 1. Simplied diagram of a single-threshold solar
energy conversion device. Photons with energy higher than
the bandgap are thermalized to the energy of the bandgap
(dashed arrow). Photons with energy lower than the bandgap
are not absorbed (pink arrow). Energy loss to separate
charge and to prevent charge recombination and drive the
system forward to do chemical or electrical work is shown
by dash-dotted arrows.
Table I: Annual Biofuel Production and Energy Conversion Efciency by Photosynthetic
Organisms and Electrical Energy Production by a Photovoltaic Cell.
Oil Producer Fuel Production [kg/(ha
year)]
Energetic Equivalent
[kWh/(ha year)]
ECE (%)
Oil palm 3,6004,000 33,90037,700 0.160.18
Jatropha 2,1002,800 19,80026,400 0.090.13
Tung oil tree
(China)
1,8002,700 17,00025,500 0.080.12
Sugarcane 2,450 16,000 0.08
Castor oil plant 1,2002,000 11,30018,900 0.050.09
Cassava 1,020 6,600 0.03
Microalgae 91,000 956,000 4.6
Si-based PV
cell
3 10
6
14.3
Source: References 14,15.
Note: ECE (energy conversion efciency) is calculated by dividing the energetic equivalent by the energy content of the
total solar spectrum averaged over 1 year incident on 1 hectare (ha) for a sunny climate at moderate latitude [21 million
kWh/(ha year)] (Phoenix, AZ; http://rredc.nrel.gov/solar/old_data/nsrdb/redbook/). Si-based PV cells produce electricity,
not fuel. We assume 80% of the hectare can be covered by solar collectors operating at 18% efciency, 0% tilt.
Comparing PV ECE with biological ECE requires knowledge of the efciency of converting either fuel to electricity or
electricity to fuel. Modern power plants can convert fuel to electricity at about 60% efciency. Electrolysis of water to
yield hydrogen and oxygen is the only commercial example of the conversion of electricity to fuel. Commercial
electrolyzers can operate at 80% efciency.
385
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
synthesis to work well with this legacy biochemistry, there
are apparent ineIfciencies; evolution might not have explored
fundamentally new pathways.
Several examples illustrate the energetic mismatch between
the add-on module, photosynthesis, and legacy bioenergetics.
The main chemical and electrochemical fuels of life [e.g., ATP
and NAD(P)H] evolved under anaerobic conditions where an
organism had to eke out a living on small redox gradients (like
running a machine with a very low voltage battery).
Photosynthesis made available both large redox gradients, span-
ning about 2.4 V in water-oxidizing photosynthesis, and highly
energetic chemical species. This thermodynamic mismatch is
apparent in Photosystem I, one of the major photosynthetic
machines of cyanobacterial and higher-plant photosynthesis,
where approximately 1.1 V (about 60% of the converted energy)
is lost before the system comes into electrochemical steady state
with the reduction of NAD(P)
+
to NAD(P)H. Even considering
that electron transfer from water to NAD(P)H also stores some
energy in ATP, the system operates at a very large overpoten-
tial. The energy loss can be approximated by comparing the
absorbed solar energy with the measured potentials of NAD(P)H
and ATP and amounts to approximately 58% of the energy enter-
ing the photochemical reaction centers [see arrow labeled
NAD(P)H + ATP in Figure 2].
Another example oI apparent ineIfciency is the water-oxi-
dizing Photosystem II, the photosynthetic reaction center that
makes the molecular oxygen that we breathe. This system is a
marvel oI thermodynamic eIfciency in the sense that minimal
overpotential appears to be necessary to oxidize watera Ieat
unmatched by any human-engineered catalystsand only
earth-abundant elements are used. However, even with elabo-
rate control mechanisms, parts of the machine break down
within 1 h under bright sunlight and must be removed from ser-
vice for repair and then reassembled.
Energy stored in ATP and NAD(P)H is used mainly to take
up and convert CO
2
into carbohydrates (also known as 'fxing
CO
2
) through a patchwork of ancient biochemical pathways
known as the Calvin cycle (or C
3
cycle). The process itself can
run on a relatively low driving force (low overpotential), but
suffers from wasteful side reactions and low maximum rates.
The key enzyme in this pathway, RuBisCO, not only catalyzes
the reduction of CO
2
to high-energy compounds, but also reacts
with O
2
, acting as an oxygenase and leading to the 'unfxing
of reduced carbon back to CO
2
and the concomitant loss of
energy in the process. In addition, RuBisCo is a notoriously
slow enzyme, with a low aIfnity to CO
2
, and thus represents a
major rate limitation for the entire photosynthetic process,
especially under conditions where CO
2
is limited. The legacy
biochemistry of the Calvin cycle is simply unable to cope with
the large infux oI solar energy available Irom photosynthesis.
Rather than solving the problem oI ineIfcient carbon fxation,
C
3
plants have evolved control mechanisms to dissipate as heat
much of the light energy they absorb under conditions where
CO
2
availability limits photosynthesis. Such dissipation is nec-
essary because, in addition to driving down the CO
2
concentra-
tion and thereby increasing the oxygenase activity, excess light
can lead to undesired chemistry (e.g., formation of reactive
oxygen species). Under full sunlight, this energy dissipation
can lead to a 5080 lowering oI the quantum yield (and there-
fore of the ECE) of photosynthesis (arrow labeled Regulatory
dissipation in Figure 2).
Some plants, algae, and cyanobacteria have partially
responded to the unfortunate situation involving RuBisCO with
a metabolically expensive patch, in which photosynthetic
energy is used to concentrate CO
2
so that it can be processed
more eIfciently by RuBisCO; this alternative provides some
improvement but is not found in all agriculturally important
plants.
Finally, as mentioned earlier, there are losses due to the
energy needed for growth and maintenance of the plant, result-
ing in a typical ECE Ior land plants oI 0.30.8 (see Table I,
which refers to biofuel production only, and Reference 4).
Interestingly, some algae and cyanobacteria (collectively
known as microalgae) have remarkably higher ECEs than
terrestrial plants (Table I). This is because some of the ECE-
limiting factors discussed above are more important for large
land plants than they are for microalgae in aqueous environ-
ments. For example, CO
2
fxation in leaves is oIten limited by
CO
2
diffusion, whereas uptake of inorganic carbon is much less
limiting in small organisms with much larger surface/volume
ratios. Moreover, because the solubility in water of CO
2
is much
higher than that of O
2
, the carbon fxation/carbon oxidation ratio
of RuBisCO is much more favorable in aqueous environments.
In addition, smaller organisms require less energy investment
for chemical transport and for generation of a skeleton, and
they tend, overall, to have better ECEs than terrestrial plants. As
shown in Table I, a far better ECE can be achieved by culturing
microalgae for biofuel production than by growing terrestrial
plants for this purpose.
Indeed, understanding the factors that limit ECE could
lead to genetic engineering efforts that enhance biofuel pro-
duction in living systems. It can be seen from Figure 2 that
regulatory dissipation is the largest loss mechanism. The need
for regulatory dissipation arises in large part from the limit-
ed kinetic competency of CO
2
assimilation. Engineering
improved kinetics of CO
2
assimilation can thus reduce this
loss and, further downstream, the loss of energy in plant
Figure 2. Simplied energy ux diagram for a typical terrestrial (crop)
plant leaf under full sunlight, showing the losses to energy conversion
efciency (ECE) during major processes of photosynthesis. The losses
are shown roughly in the temporal sequence in which they occur; the
width of the arrows is qualitatively proportional to the percentage of
energy lost. Internal conversion refers to the relaxation to the threshold
indicated by the dashed arrow in Figure 1. Feedback from the three
metabolic processes and the incident sunlight intensity strongly affects
the magnitude of regulatory dissipation.
386
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
growth and maintenance. Much more needs to be understood
about photosynthesis (and plant metabolism in general) to
make larger gains, perhaps involving radical reengineering of
plants to decrease the loss of energy in coupling the photo-
chemical events to NAD(P)H and ATP synthesis.
How Can Articial Photosynthesis Be Used To
Address These Issues?
EIfcient water oxidation is one oI the holy grails oI sustain-
able solar fuel production. As mentioned above, water oxida-
tion in photosynthesis is essentially a single-threshold process
and is thereIore subject to the ShockleyQuiesser limit oI about
30% ECE. This value is further reduced to about 10% by the
use of the two photosystems (two quanta per electron), each
with nonoptimal thresholds of about 1.8 eV. It has not been
demonstrated that this tandem system is necessary to split
water; it is almost certainly a legacy of the evolution of type 1
(Photosystem I, PS I) and type 2 (Photosystem II, PS II) reac-
tion centers. To explore the two-quanta-per-electron issue, arti-
fcial reaction centers are being used in conjunction with
dye-sensitized semiconductor photoelectrodes (Grtzel-type
photoanodes) to determine the overpotentials necessary to drive
water oxidation and H
2
production at rates comparable to the
solar photon fux, using biological and bio-inspired catalysts.
Once the electrochemical parameters that optimize single-pho-
ton-per-electron water oxidation are determined, efforts already
underway to reengineer photosynthetic membranes will be bet-
ter informed.
Reengineered photosynthesis could involve pathways that
are not limited by legacy biochemistry. For example, only three
enzymatic steps, all catalyzed by NAD-linked dehydrogenases,
are required to oxidize methanol to CO
2
. If remodeled to run
backward and be driven directly by PS II, which could be engi-
neered to have a slightly more reducing potential, this process
would accomplish one-photon-per-electron water oxidation
and fuel production. There would be a small penalty associated
with moving the 1.8 eV chlorophyll excited state higher (and
further from the optimal bandgap) to increase the reduction
potential, but that would be more than offset by achieving one-
photon-per-electron photochemistry.
Artifcial photosynthetic constructs and bio-inspired
catalysts can serve as links between human-engineered and
biological processes and thereby open the door to new carbon-
fxing pathways. In the example above, the remodeled NAD-
linked enzymes driven by sustainably produced emf
(electricity) could direct the reduction of CO
2
to methanol. If
methane mono-oxygenase were added and the reaction driven
backward, the product would be the worlds most energy-rich
carbon-based compound (on a per-carbon basis): methane.
The advantages of enzymes (or, someday, bio-inspired cata-
lysts) that oxidize water and direct the synthesis of energy-
rich Iuel at room temperatureand that require only electrical
driving Iorce and redox equivalentsset up an overarching
challenge in this feld: coupling the appropriate redox enzymes
to sources oI emI. Toward this end, artifcial photosynthetic
reaction centers have been demonstrated to photo-inject elec-
trons or holes into semiconductors. These processes convert
molecular redox potentials of the kind used by redox enzymes
to emI in metallic conductors. This feld is advancing rapidly.
Hoped-for breakthroughs include the design and synthesis of
new high-potential electron mediators inspired by the tyro-
sine/histidine system found in water-oxidizing photosynthesis
and the design, synthesis, and assembly oI artifcial active
sites of enzymes directly on electrodes. The development of
ultrahigh-surface-area materials that could provide the three-
dimensional structure necessary to recognize transition
states, and thereby guide reactants to desired products, and
also conduct emf to drive the reactions would make available
eIfcient electricity-to-Iuel technology. Corrosion-resistant
semiconductors having valence-band holes at the potential
necessary to oxidize water and conduction bands that are elec-
trochemically negative enough to reduce protons to hydrogen
(hydrogen production is a thermodynamic proxy for the
synthesis of all energy-rich, carbon-based fuels) present a sig-
nifcant challenge to materials science but would enable solar
water splitting.
Such work illustrates the interplay between natural and arti-
fcial constructs that could lead to new materials or compounds.
For example, the ECE of the Grtzel PV cell would be substan-
tially improved by a high-potential electron relay mimicking
the tyrosine/histidine-based system used by nature in PS II.
Energy Conversion by Hybrids of Natural and
Articial Systems
Once effective molecular wires or other electron relays
between metallic conductors and genetically engineered bio-
logical catalysts can be made, microorganisms dedicated to
synthesis could provide a renewable source of catalytic mate-
rial. In this way, the self-assembly, self-repair, and self-replica-
tion characteristics of living things would be harnessed to meet
human needs.
Recent discoveries indicate that nature might, in fact, have
already outlined a strategy for the exchange of energy in the
form of emf between human-engineered sources and biochemi-
cal processes. Geobacteria appear to produce pili that make
electrical connections across their plasma membrane, connect-
ing the redox chemistry inside the organism to an external elec-
tron sink. Imagine that, when coupled to an electrode with
appropriate wiring to complete the circuit, an applied potential
could drive metabolic processes inside the organism. Such a
system in which the organism was genetically reengineered to
produce biofuel would be self-perpetuating in the biological
sense but inexorably linked to human-engineered energy
sources that are eIfcient and robusta true hybrid.
In addition to providing catalytic paradigms for much of the
mainstream energy-processing chemistry that humans use,
nature almost certainly has found catalytic pathways for other
chemical processes that are key to the eIfcient and sustainable
use oI Earth`s resources. Nitrogen fxation is one obvious exam-
ple. Another is digestion of cellulose to simple sugars. Powerful
new tools such as metagenomics are being used to discover new
organisms that have metabolic pathways tuned to match their
local niches. Indeed, we have undoubtedly only just begun to
learn from nature.
For Iurther inIormation on engineered and artifcial photo-
synthesis, the interested reader can consult ReIerences 113.
References
1. D. Gust, T.A. Moore, A.L. Moore, Acc. Chem. Pes. 34, 40 (2001).
2. D. Gust, T. Moore, A. Moore, in Arr|/|c|a| Phorosynrhes|s, A.F. Collings,
C. Critchley, Eds. (Wiley-VCH, Weinheim, Germany, 2005), p. 187.
3. M. Hambourger, P. Liddell, D. Gust, A.L. Moore, T.A. Moore, Phorochem.
Phorob|o|. Sc|. 6, 431 (2007).
4. S.P. Long, X.-G. Zhu, S.L. Naidu, D.R. Ort, P|anr Ce|| Env|ron. 29, 315
(2006).
5. G.W. Huber, S. Iborra, A. Corma, Chem. Pev. 106, 4044 (2006).
6. A.J. Bard, M.A. Fox, Acc. Chem. Pes. 28, 141 (1995).
7. M.D. Archer, J.R. Bolton, J. Phys. Chem. 94, 8028 (1990).
8. D.A. Bryant, N.-U. Frigaard, 7rends M|crob|o|. 14, 488 (2006).
9. D.R. Lovley, Nar. Pev. M|crob|o|. 4, 497 (2006).
387
||'J||||'

||JM/'' & ||J||||'
-23 "5,,%4).

\J||M|

/|||| I11

WWWMRSORGBULLETIN

|a|a-..|a Ma|-||a|. |:| |a-|
10. M. Grtzel, lnorg. Chem. 44, 6841 (2005).
11. Y. Chisti, B|orechno|. Adv. 25, 294 (2007).
12. T. Searchinger, R. Heimlich, R.A. Houghton, F. Dong, A. Elobeid, J.
Fabiosa, S. Tokgoz, D. Hayes, T.-H. Yu, Sc|ence, February 7, 2008
(10.1126/science.1151861).
13. J. Fargione, J. Hill, D. Tilman, S. Polasky, P. Hawthorne, Sc|ence,
February 7, 2008 (10.1126/science.1152747).
14. J.A. Duke, Handbook of Energy Crops (1983), www.hort.purdue.edu/
newcrop/duke_energy/dukeindex.html (accessed January 2008).
15. O. Pulz, Performance Summary Report: Evaluation of GreenFuels 3D
Matrix Algae Growth Engineering Scale Unit (APS Red Hawk Power Plant,
AZ, JuneJuly 2007), www.greenfuelonline.com/gf_les/Performance%20S
ummary%20Report.pdf (accessed January 2008). N